Not so long ago, it was very likely that those of us who lived through the early eras of IT and networking in companies would have had to face and search for thousands of explanations about the downtime of company applications and servers.
The truth is that from the years of that beloved "Quartico" to today's Data Centers, we have gone through a multitude of transformations, changes and improvements. But believe it or not -even for many companies-, the question we mentioned above, nuanced and updated to current times, is still valid.
For many areas of IT, the needs remain more or less the same, and very little has changed. The big difference is that some of those responsibilities have had to be transferred to technology services companies that, in particular, have much more experience, resources and can keep up with the pace of upgrades and changes.
Coincidentally, the most obvious thing that should be outsourced, such as the data center, is often the last thing on the list.
So what is a data center? It is, in essence, everything we used to have in the "Quartic" under the stairs -and much more-, but operating in a room with all the physical and environmental safeguards. The reason? The same reason why we no longer keep our money under the mattress like our grandmothers did: security.
And safety involves a lot of things. We often don't know how costly a failure can be until the most insignificant communications equipment stops working. In those days of yore, it was normal for something to go down because the janitor hung the lunch box from the breaker that powered the computer room. It used to be possible that we could even cool our computers with a home mini-split. It was also quite normal that in the same "Cuartico" several of the cartons with which the equipment was packed were kept. Do not ask where the nearest fire extinguisher was.
In short, we can never separate the word "security" from the word "risk" and nothing could be truer than in the context of a Data Center: nowadays data is the most important strategic asset of any organization and we can't take the chance that someone, by mistake, formats an entire server that we don't back up because the backup tapes have run out.
Words more, words less, the Data Center is the big and rich cousin of the "Quartic" of yesteryear, only that nowadays it is not possible to conceive the Data Center only as that place that houses the network of computer equipment of an organization, which includes the power supply, air conditioning systems and security applications. The days of renting space as if they were rooms in a hostel still exist and are, to some extent, necessary. But they are offers with very basic characteristics, little differentiated and where basically "the war of the penny" is lived.
Data Center service providers know very well that the real monetization and value for customers comes from a whole portfolio of complementary services offered inside and outside the Data Center ecosystem, but that thanks to connectivity and IP networks it is possible to integrate them with customers' co-location infrastructures.
While it is feasible today to continue with the on-premise Data Center model, meeting many of the characteristics and conditions of a Data Center as a service, the question again is whether the "with almost everything" covers the risks that the organization is willing to tolerate, accept and manage. Not to speak, obviously, of the CAPEX and OPEX budget that for an organization implies to implement, operate and maintain even the most modest Data Center.
In that sense, today's market offers a variety of options and typologies to suit all requirements and budgets, which we have conveniently summarized in the following illustration:
The term TIER alludes to the way Data Centers are stratified or classified according to the redundancy levels of some of their components, especially the electrical ones.
The first level or TIER 1 refers to a basic, dedicated infrastructure with practically no redundancy, which means that there is a probability of having outages or unavailability equivalent to 28 hours per year. These types of options are ideal for small businesses that want to make the "quantum leap" from the beloved "Quantum" to something much, much better. It is the equivalent of going from ox cart to 4-stroke car in the same day. The IT organization and operation will have a hard time understanding this, but it's a huge improvement in every way. The bad news is that the supply of this type of Data Centers is practically non-existent in our market. Market opportunity? Not so much. We'll get to it.
The TIER 2 Data Center has a redundant infrastructure, which although it reduces the probability of downtime to only 22 hours per year, the truth is that the level of redundancy is basic and not on all elements. In practical terms, it is like having the servers of the "Quartic" connected to two different multisockets, even though both are connected to the same circuit in the breaker box. Is this an improvement? Yes, of course, but it is often not enough. Although there is a supply of this type on the market, the reality is that they are scarce.
TIER 3 has almost complete fault tolerance, since it offers at least 72 hours of protection in the absence of power in the Data Center circuits. The downtime is reduced to a comfortable 1.6 hours per year, which (for the vast majority of large companies) is far more than sufficient. Keeping the distance, it is as if you were to put a couple of electric generators in your "Cuartico" of the old days, which should be ready to go into action if the power goes out. This is the most abundant offer in the market and targets a wide variety of sectors, companies and markets.
Finally, TIER 4 provides total and complete fault tolerance at various levels. Here the concept of redundancy is not only conceived in terms of cooling components, electrical system, fire suppression system and security access, but provides multiple paths, multiple sources, with several separate geographical locations. This is particularly important with external circuits feeding the data center. This scenario is difficult to replicate on premise, even for large organizations, as the level of investment, operation and implementation required far exceeds the needs of the business. There is demand and access to this type of data centers in the market, but their contracting is practically limited to large corporations, carriers, banks, some government companies and multinationals, which are the ones with the budget and the sufficiently justified need to be able to access this type of offer.
And how does the cloud fit into all this? It is a complement. Until a couple of years ago, the paradigm of the Cloud, especially the public one, was based on a practically linear path: the Journey to Cloud 1.0, where it was intended that the customer would go through a series of stages as if it were a multilevel system, where only those who managed to migrate all their workloads, servers and applications to public clouds, would be the ones who could access the best benefits and possibilities in the technological field.
But this journey soon ran up against the reality of the vast majority of customers: the digital debt of the non-migratable, the investment of bringing data onshore and the unpredictable recurring costs. In this context the journey brought "where it suits me best" into the equation and the concept of the Hybrid Cloud emerged.
The Data Center -on premises or as a service- is part of this Hybrid Cloud concept, since, in this heterogeneous scenario, services can be composed of loads or subservices that are in different layers and all working in an integrated manner.
Finally, much water has flowed since the days of the "Quartic". Today, data centers represent a big step forward in terms of security, reliability and availability of IT services. However, for those who still harbor fears about making this leap, the good news is that with the advent of the Hybrid Cloud, taking this step does not necessarily have to be a Big-Bang.
It can happen in a phased manner, in phases and transitions that best suit the needs and realities of organizations. The Data Center is a stepping stone, a lever that helps to support a framework that needs to be sustained at various points and ensure that it is stable, secure and reliable.
Much of the success of data centers is based on their ability to interconnect in a reliable and cost-efficient way to the global ecosystem.